Behavioral marital therapy (BMT) with alcoholics produces better outcomes in the two years after treatment than methods that do not involve the spouse or other family members. However outcomes vary considerably. Some couples get substantial and enduring benefits and some suffer serious relapses. Further, BMT research has included few women and no minorities. The proposed grant will oversample women and minority alcoholics to determine whether positive results of BMT studies extend to more heterogeneous "real world" patient populations. Adding the new subjects to an ongoing NIAAA project will provide a sample of 400 couples (100 male and 100 female minority plus 100 male and 100 female non-minority alcoholics) treated in BMT with outcome data collected quarterly for a two year follow-up period. The primary aim of the overall project is (a) to identify pretreatment characteristics of alcoholic couples that predict response to BMT and that discriminate successful from unsuccessful outcomes after BMT and (b)to determine whether predictors and patterns of response to BMT differ as a function of gender and minority status. Analyses pertinent to the primary aim of this research will address three interrelated sets of questions that are divided into three substudies for clarity. In each substudy, analyses will examine whether predictors and patterns of response to BMT differ as a function of gender and minority status. Study 1 will examine alcoholic patient characteristics, marital relationship, and other factors as possible predictors of outcome after BMT. Prior research has not examined factors that predict response to BMT (or any other type of marital/family therapy). Patient characteristics have been studied extensively as predictors of response to alcoholism treatment, while marital relationship and other factors have not. Therefore, this study will examine specifically the extent to which marital and other factors contribute over and above patient characteristics. Unlike the alcoholism literature, studies relating marital/family factors to treatment outcomes for other types of psychopathology (schizophrenia, depression) have a long history. A well replicated finding is the positive relationship between high levels of negative expressed emotion (EE) by relatives about the patient and subsequent relapse. No work has been published on EE and alcoholism. Study 2 will examine whether high levels of spouses' EE at intake predicts relapse among alcoholics in the two years after BMT. Clinicians and program planners need to know whether or not an individual patient is likely to benefit from a given treatment method. Study 3 will examine what proportion of cases treated in BMT obtain clinically significant improvements in drinking and marital adjustment and which factors discriminate cases which obtain successful outcomes from those that do not. Extent of reductions in domestic violence after BMT also will be considered. A secondary aim, pursued in Study 4, is to develop and test a longitudinal treatment process model that examines whether the extent of couples' use of BMT-targeted behaviors (and other methods such as AA) during and after treatment explains outcomes after BMT more fully than pretreatment characteristics alone.